Edinburgh Research Explorer
Genetic correlations among protein yield, productive life, and
type traits from the United States and diseases other than
mastitis from Denmark and Sweden
Citation for published version:
Rogers, GW, Banos, G & Sander-Nielsen, U 1999, 'Genetic correlations among protein yield, productive life,
and type traits from the United States and diseases other than mastitis from Denmark and Sweden' Journal
of Dairy Science, vol 82, no. 6, pp. 1331-1338. DOI: 10.3168/jds.S0022-0302(99)75357-9
Digital Object Identifier (DOI):
10.3168/jds.S0022-0302(99)75357-9
Link:
Link to publication record in Edinburgh Research Explorer
Document Version:
Publisher's PDF, also known as Version of record
Published In:
Journal of Dairy Science
General rights
Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s)
and / or other copyright owners and it is a condition of accessing these publications that users recognise and
abide by the legal requirements associated with these rights.
Take down policy
The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer
content complies with UK legislation. If you believe that the public display of this file breaches copyright please
contact [email protected] providing details, and we will remove access to the work immediately and
investigate your claim.
Download date: 15. Jun. 2017
Genetic Correlations Among Protein Yield, Productive Life,
and Type Traits from the United States
and Diseases Other than Mastitis from Denmark and Sweden
G. W. ROGERS,*,1 G. BANOS,† and U. SANDER-NIELSEN‡
*Department of Dairy and Animal Science, The Pennsylvania
State University, 324 Henning Building, University Park 16802
†INTERBULL Centre, Dept. of Animal Breeding and Genetics,
Swedish University of Agricultural Sciences,
Box 7023, Uppsala, Sweden S 750 07
‡Danish Agricultural Advisory Center,
Udkaersdej 15, Aarhus, Denmark 8200
ABSTRACT
Sire genetic evaluations for protein yield, productive life, and selected type traits from the US were
correlated with sire evaluations for disease from Denmark and Sweden and were then adjusted to approximate genetic correlations. Disease categories from
Denmark included reproductive diseases, foot and leg
diseases, metabolic and digestive diseases, and all
diseases other than mastitis. Genetic evaluations for
Denmark were from separate analyses for each disease category using a multiple-trait sire model with
first, second, and third lactations handled as multiple
traits. Evaluations from Sweden for all diseases other
than mastitis were from a single-trait sire model
using only first lactations. In addition, Danish and
Swedish genetic evaluations were regressed on US
type evaluations to test for quadratic relationships.
Relationships were based on 104 bulls with US and
Danish evaluations (88 with US type) and 84 bulls
with US and Swedish evaluations (83 with US type).
Genetic correlations between US protein yield and
diseases were unfavorable, but correlations were
favorable between productive life and disease.
Genetic correlations among US type and diseases
were around zero, except for correlations with US
dairy form (range –0.34 to –0.73). Genetic correlations calculated from residual correlations (adjusted
for predicted transmitting abilities for milk) between
productive life and diseases were favorable (range
0.29 to 0.51). Genetic correlations calculated from
residual correlations (adjusted for predicted transmitting abilities for milk) between dairy form and
diseases ranged from –0.10 to –0.53. Selection for
increased productive life may reduce disease occur-
Received July 27, 1998.
Accepted February 5, 1999.
1Work completed while this author was on sabbatical at INTERBULL Centre in Uppsala, Sweden.
1999 J Dairy Sci 82:1331–1338
rences, but selection for higher dairy form scores will
increase disease occurrences.
( Key words: diseases, yield, type traits, longevity)
Abbreviation key: PL = productive life.
INTRODUCTION
It is widely accepted (7, 10) that a genetic antagonism exists between milk yield in dairy cattle and
disease incidence. However, selection for improved
yield, especially protein yield, is economically justified ( 7 ) . Likewise, disease incidence and severity
warrants selection pressure if accurate recording is
possible at a reasonable cost. When direct recording of
disease incidence is difficult, other measures may be
useful for indirect selection to improve disease
resistance or at least moderate the undesirable
response of disease that occurs with intense selection
for yield (6, 7, 8, 9, 10).
Some possible traits that could be useful for indirect selection of improved disease resistance include
measures of longevity such as productive life ( PL)
and measures of physical characteristics such as body
weight, body condition, or linear type traits that
reflect body structure. Selection for increased PL
could reduce diseases because diseases are primary
causes of premature culling. Selection for increased
body condition or less change in body condition during
the lactation may reduce metabolic diseases in dairy
cattle. Veerkamp ( 1 3 ) outlined the phenotypic associations among metabolic and reproductive problems
and various measures of energy balance. Negative
energy balance had a phenotypic association with
several disorders. Some authors (4, 13) have suggested that physical traits including body weight
could be used to select for improved feed efficiency.
Selection to reduce body weight in an effort to improve feed efficiency is justified unless selection to
reduce body weight results in substantial increases in
1331
1332
ROGERS ET AL.
disease occurrences. Selection for smaller body weight
or size (along with selection to increase yield) may
lead to more metabolic stress and may increase the
frequency of metabolic and reproductive diseases. Any
results of simultaneous selection for increased yield
and smaller body size would depend on the genetic
correlations among body size, yield, and diseases,
which are largely unreported (13). Indeed, estimates
of genetic relationships among PL, body dimensions,
type traits, and diseases are needed to properly utilize these traits in breeding programs.
Currently in the US and some other countries,
some selection occurs for increased body size and
increased angularity or dairy character. This selection
is due to the emphasis placed on selection for increased yield but also is due to selection based on type
traits. Selection for higher type scores essentially
places a small direct emphasis on increased body size
and increased angularity. Perhaps more important,
angularity or dairy character has been used to help
identify dams of bulls destined for progeny test programs.
The objectives of this paper are to estimate genetic
relationships among protein yield, PL, and type traits
that reflect body and locomotive characteristics from
the US and diseases other than mastitis from Denmark and Sweden. The objectives include characterization of quadratic relationships among genetic
evaluations for diseases other than mastitis and
linear type traits that reflect body and locomotive
characteristics.
MATERIALS AND METHODS
Official sire evaluations from the US (July 1995)
for production traits, PL, and type traits (from
USDA-DHIA, Beltsville, MD and Holstein Association, Brattleboro, VT) were used in estimating the
genetic correlations. Details of the US genetic evaluations can be found elsewhere (11, 12, 14). Traits were
summarized using animal models with heritabilities
ranging from 0.085 (for PL) to 0.42 (for stature).
Heritability for protein yield was 0.25.
Official sire evaluations from Sweden for diseases
other than mastitis (July 1995) were also used in the
analyses. Genetic evaluations from Sweden include
information from all diseases other than mastitis
(composite trait) analyzed as a binomial trait.
Veterinary treatments and fertility treatments by artificial insemination technicians from 10 d before first
calving to 150 d after first calving were considered.
The evaluations were calculated using a single-trait
sire model with relationships (sire and maternal
grandsire) and a heritability of 0.02 ( 3 ) .
Journal of Dairy Science Vol. 82, No. 6, 1999
Unofficial sire evaluations from Denmark for diseases other than mastitis were also used in the analyses. Danish evaluations were calculated with data
for Danish Black and White cattle from the Danish
health recording system using a multiple-trait sire
model that included lactations 1, 2, and 3 as separate
traits. The model included the effects of herd and year
of calving, year and month of calving, calving age,
heterosis (to account for crossing North American
Holsteins and Danish Friesians), sire (random), and
residual. Relationships among sires through sire
paths were included. Sire evaluations (from separate
analyses) were available based on all diseases other
than mastitis (composite trait, defined similarly to
the Swedish trait), reproductive diseases, feet and leg
diseases, and digestive and metabolic diseases.
Heritabilities were all around 0.02, and genetic correlations between the lactations for each disease trait
were large. Disease incidence was defined as 1, 2, or
≥ 3 episodes in a lactation. The lactation period was
defined as 10 d before calving until d 305 of lactation.
Danish evaluations from the analyses were multiplied
by 10 and reversed in sign before use in computation
of correlations and regressions. As a consequence,
higher values for Danish genetic evaluations were
favorable and reflected a lower disease frequency,
which was also true for the Swedish evaluations.
More details on the data, genetic parameters, and the
method for calculating the genetic evaluations can be
found in Sander-Nielsen et al. ( 9 ) .
Sire genetic evaluations from the US were merged
with sire genetic evaluations from Denmark to establish a file that included bulls with evaluations and
daughters in both countries (i.e., US-Denmark file).
An international cross-reference file established by
the INTERBULL Centre (Uppsala, Sweden) was
used to facilitate file merges. The same procedure was
used to establish a file that included bulls with evaluations and daughters in Sweden and the US (i.e., USSweden file). Genetic correlations were estimated by
adjusting product-moment correlations among sire
evaluations for reliabilities ( 1 ) . Each genetic correlation was calculated by dividing the product-moment
correlation among sire evaluations for the two traits
by the square root of the product of the mean reliabilities for the two traits involved. The correlation between daughter deviations on two traits, for which
one trait is measured for one group of daughters and
the second trait is measured for a second group of
daughters sired by the same bulls, is an estimate of
the genetic correlation. The method of Calo et al. ( 1 )
uses this concept but relies on genetic evaluations
instead of daughter deviations. The method
eliminates the impact of having genetic evaluations
1333
TRAITS CORRELATED WITH HEALTH
regressed toward the mean (compared with daughter
deviations) when the reliability of the genetic evaluations is less than 1.0. Residual correlations after adjusting for birth-year of the bull to eliminate the
effect of genetic trend on correlation estimates were
also calculated but not reported because they were
very similar to the product-moment correlations.
Genetic trend in the traits can bias genetic correlations estimated using genetic evaluations from independent daughter groups. Adjustment for birth
year would likely eliminate the impact of genetic
trend and any potential bias in the correlations that
might result from genetic trend. Residual correlations
among Danish and Swedish traits and US PL and
dairy form after adjusting for PTA milk yield were
also calculated, and genetic correlations calculated
from these residual correlations are reported. In addition, genetic evaluations for disease data from Denmark and Sweden were regressed in separate models
on US type traits to test for linear and quadratic
relationships.
Genetic evaluations from the US and from Denmark or Sweden are from independent daughter
groups, so only genetic covariance should be responsible for the correlations among progeny group performance. Edits were made to include only sires with 50
daughter equivalents for all disease traits in Denmark and reliability for PL from the US of 0.60 or
greater or, in the case of matches with type, reliability for linear type of 0.70 or greater when creating the
US-Denmark file. Edits were made to include only
sires with 50 daughter equivalents in Sweden and
reliability for PL from the US of 0.60 or greater or, in
the case of matches with type, reliability for linear
type of 0.70 or greater when creating the US-Sweden
file. The same data were used for the regression
analyses. Genetic correlations calculated from evaluations on sires with 125 daughter equivalents or more
in Denmark or Sweden were made but not reported
because these genetic correlations were very similar
to those calculated from evaluations on sires with 50
daughter equivalents or more. Correlations with diseases in the third lactation from Denmark were also
not reported because they were almost identical to the
correlations involving diseases in second lactation.
Genetic correlations between diseases in second and
third lactations in Denmark are very high ( 9 ) . One
undesirable characteristic of the estimation procedure
is the potential to get estimates outside the
parameter space due to the adjustment for reliabilities, which can accentuate sampling effects. This
problem diminishes as reliabilities of the evaluations
increase. At the limit for reliabilities, the correlations
among genetic evaluations from separate data sources
represent the estimated genetic correlation. Stability
of the estimates across various edits increases the
confidence associated with the estimates. Utilizing
genetic evaluations from independent daughter
groups to calculate genetic correlations allows one to
approximate partial genetic correlations (genetic
correlations after the removal of the genetic contribution of another trait) and to calculate genetic regressions.
RESULTS AND DISCUSSION
Means, standard deviations, and descriptions for
the sire evaluations used in the study are in Table 1.
For the Danish and Swedish traits, higher sire evaluations are more desirable. Mean reliabilities for the
US traits were all 0.95 or greater. Mean reliabilities
for the Danish evaluations were from 0.49 to 0.56 and
depended on the trait and data subset (matched with
US yield data or type data). Mean reliability for the
Swedish trait was 0.51. Adjustments to productmoment correlations to obtain genetic correlations are
dependent on these reliabilities. Product-moment
correlations among the genetic evaluations can be
calculated by multiplying the estimated genetic correlations by a factor that depends on the reliabilities of
the traits involved. Product-moment correlations can
be calculated by multiplying the genetic correlations
by a factor that ranges from 0.74 to 0.69 in the USDenmark file and by approximately 0.70 in the USSweden file. Adjustments to product-moment correlations to obtain genetic correlations were primarily a
reflection of the mean reliabilities of the Danish and
Swedish evaluations because the mean reliabilities
for all US evaluations were very high.
Genetic correlations and genetic relationships calculated primarily from information on daughters of
selected bulls may or may not be representative of the
true genetic correlation in the population. However,
the genetic relationships calculated using information
from daughters of selected bulls is representative of
the genetic relationships within the contemporary
breeding population. In addition, it should be noted
that the genetic correlations used to calculate acrosscountry evaluations and conversions come primarily
from information on selected bulls and their close
relatives (similar subset of bulls used in this study).
Genetic correlations among PL and protein yield
from the US and the composite trait that includes all
diseases other than mastitis from Denmark and
Sweden are in Table 2. Genetic correlations between
protein yield and the composite disease traits were
from –0.19 to –0.62 and unfavorable. These correlations agree in direction with other studies (5, 7, 9,
Journal of Dairy Science Vol. 82, No. 6, 1999
1334
ROGERS ET AL.
TABLE 1. Mean, standard deviation, and description for sire genetic evaluations from the US,
Denmark, and Sweden.1
Traits
US
Protein yield, kg
Productive life, mo
Final score
Rear legs side view
Foot angle
Rump angle
Rump width
Dairy form
Body depth
Strength
Stature
Danish2
All diseases other than mastitis
First lactation
Second lactation
Reproductive diseases
First lactation
Second lactation
Foot and leg diseases
First lactation
Second lactation
Digestive and metabolic diseases
First lactation
Second lactation
Swedish3
All diseases other than
mastitis in first lactation
X
SD
Higher values for the
trait correspond to
1.65
0.497
0.397
–0.135
0.302
–0.244
0.235
0.441
0.289
0.184
0.252
10.4
1.32
0.834
1.27
1.38
1.19
1.33
1.08
1.32
1.33
1.24
Higher yield
Longer life
Higher final score
More leg set
Steeper angle
Lower pin setting
Wider rump
More angularity
More depth
More strength
More height
–0.600
–0.594
0.394
0.370
Less disease
Less disease
–0.202
–0.201
0.168
0.186
Less disease
Less disease
–0.088
–0.066
0.112
0.080
Less disease
Less disease
–0.255
–0.271
0.195
0.200
Less disease
Less disease
94.8
4.68
Less disease
1Data
for US evaluations of protein yield and productive life and Danish evaluations of diseases are
from 104 sires. Data from US evaluations of type are from 88 sires. Data from Swedish evaluations of
diseases are from 84 sires.
2Disease incidence for each trait definition and lactation were scored 1, 2, or 3 ( 3 and greater were
combined). Danish evaluations were multiplied by –10 to scale the evaluations to be equivalent to
published evaluations (higher values favorable) in Denmark and Sweden.
3Standardized to a mean of 100 and standard deviation of five for population.
10), however, the correlations between protein yield
and the Danish measures were higher than in most
reports. The correlations between protein yield and
the composite disease traits in Denmark were larger
in magnitude than between protein yield and the
composite disease trait from Sweden. The differing
results could be due to sampling or variation in
recording and sire evaluation methods. Of the 104
and 84 bulls in the US-Denmark and US-Sweden
files, only 39 were common to both the US-Denmark
and US-Sweden files. Correlations in this study with
milk and fat yield (not reported) were similar to the
correlations with protein yield.
Genetic correlations between PL and composite disease traits in Table 2 were favorable and ranged from
0.16 to 0.39. The PL evaluations from the US had
very high reliabilities (mean > 0.95), so essentially
all of the information in the PL evaluations would
Journal of Dairy Science Vol. 82, No. 6, 1999
have been direct information on daughter PL (PL
evaluations on bulls in this study would have little
influence from the correlated traits used in calculating US PL evaluations). Correlations between PL on
bulls with early first-crop progeny and disease might
not be the same as those reported here because PL
evaluations on newly evaluated bulls are impacted by
type traits used to predict PL. Many of these type
traits may have little predictive value for diseases
other than mastitis.
Genetic correlations between selected type traits
from the US and the composite disease traits from
Denmark and Sweden are in Table 2. Most correlations were near zero or were not consistently positive
or negative. This result might be expected because
the composite disease traits are a reflection of many
different types of diseases. One exception was found;
the correlations between dairy form and the composite disease traits were all substantial and negative
1335
TRAITS CORRELATED WITH HEALTH
TABLE 2. Approximate genetic correlations among productive life,
protein yield, and selected type traits from the US and all diseases
other than mastitis from Denmark and Sweden.1
All diseases other than mastitis
US trait
First
lactation
in Denmark
Second
lactation
in Denmark
First
lactation
in Sweden
Protein yield
Productive life
Final score
Rear legs side view
Foot angle
Rump angle
Rump width
Dairy form
Body depth
Strength
Stature
–0.62*
0.16
–0.20
–0.15
0.05
0.20
–0.18
–0.73*
–0.25
–0.14
–0.13
–0.60*
0.23
–0.06
–0.04
0.09
0.05
–0.06
–0.60*
0.00
0.06
0.09
–0.19
0.39*
0.06
–0.04
0.09
–0.14
0.20
–0.46*
0.03
0.13
0.15
1Correlations between US protein yield and productive life and
Danish traits are based on data from 104 bulls. Correlations between US type traits and Danish traits are based on data from 88
bulls. Correlations between US protein yield and productive life
and the Swedish trait are based on data from 84 bulls. Correlations
between US type traits and the Swedish trait are based on data
from 83 bulls. Edits were made to include only bulls with an
approximate minimum of 50 daughter equivalents in the genetic
evaluations for diseases and reliabilities for US productive life of
0.60 or greater or reliabilities for US linear type of 0.70 or greater.
Because of scaling, higher values are desirable for the Danish and
Swedish traits.
*Correlations among sire evaluations were different from 0 ( P <
0.05).
(range –0.46 to –0.73). Dairy form is correlated with
yield, so this result agrees with the correlations between protein yield and disease. Higher dairy form
scores were genetically associated with increased disease frequency.
Genetic correlations among US traits and
reproductive diseases, foot and leg diseases, and
digestive and metabolic diseases from Denmark are
in Table 3. Protein yield had a substantial antagonistic genetic correlation with all the disease categories.
Selection for increased protein yield will likely lead to
increased reproductive diseases, increased foot and
leg diseases, and increased digestive and metabolic
diseases. The increased stress that comes with higher
yield apparently increases the frequency of all the
major diseases in dairy cattle.
Genetic correlations among PL and reproductive
diseases, foot and leg diseases, and digestive and
metabolic diseases were all small but favorable.
Correlations with reproductive diseases tended to be
smaller than correlations with foot and leg diseases or
digestive and metabolic diseases. Selection for increased PL may lead to decreased disease frequency
in dairy cattle. Productive life in the US is affected by
yield ( 1 1 ) and has a positive correlation with yield
traits (not adjusted for yield). The small positive
correlations among PL and the disease traits exist in
spite of the undesirable contribution from higher yield
on diseases and the impact of yield on PL evaluations.
Genetic correlations among US PL and the disease
traits calculated from residual correlations after adjustment for PTA milk yield are in Table 4. Correlations between disease traits and PL are all positive
(range 0.29 to 0.51) and larger than the corresponding genetic correlations calculated from productmoment correlations among sire evaluations. The
genetic component of PL that is independent of milk
yield has a higher correlation with disease traits than
TABLE 3. Approximate genetic correlations among productive life, protein yield, and selected type traits from the US and reproductive
diseases, foot and leg diseases, and digestive and metabolic diseases from Denmark.1
Reproductive
diseases
Foot and leg
diseases
Digestive and
metabolic diseases
US trait
lactation 1
lactation 2
lactation 1
lactation 2
lactation 1
lactation 2
Protein yield
Productive life
Final score
Rear legs side view
Foot angle
Rump angle
Rump width
Dairy form
Body depth
Strength
Stature
–0.52*
0.08
–0.21
–0.14
–0.02
0.08
–0.08
–0.64*
–0.19
–0.09
–0.09
–0.62*
0.12
–0.16
–0.07
–0.06
–0.08
–0.13
–0.61*
–0.13
–0.04
–0.04
–0.47*
0.19
–0.09
–0.07
–0.01
0.39*
–0.20
–0.50*
–0.28*
–0.20
–0.13
–0.46*
0.18
–0.02
–0.02
–0.05
0.40*
–0.05
–0.38*
–0.12
–0.08
–0.02
–0.43*
0.17
–0.06
–0.11
0.11
0.16
–0.13
–0.55*
–0.11
–0.05
–0.02
–0.30*
0.24
0.06
–0.08
0.12
0.09
0.02
–0.34*
0.13
0.12
0.19
1Correlations between US protein yield and productive life and Danish traits are based on data from 104 bulls. Correlations between
US type traits and Danish traits are based on data from 88 bulls. Edits were made to include only bulls with an approximate minimum of
50 daughter equivalents in the genetic evaluations for diseases and reliabilities for US productive life of 0.60 or greater or reliabilities for
US linear type of 0.70 or greater. Because of scaling, higher values are desirable for the Danish traits.
*Correlations among sire evaluations were different from 0 ( P < 0.05).
Journal of Dairy Science Vol. 82, No. 6, 1999
1336
ROGERS ET AL.
TABLE 4. Approximate genetic correlations among productive life
and dairy form from the US and diseases other than mastitis from
Denmark and Sweden calculated from residual correlations among
sire genetic evaluations after adjustment for PTA milk yield.1
US
All diseases other than mastitis
First lactation in Sweden
First lactation in Denmark
Second lactation in Denmark
Reproductive diseases
First lactation in Denmark
Second lactation in Denmark
Foot and leg diseases
First lactation in Denmark
Second lactation in Denmark
Digestive and metabolic diseases
First lactation in Denmark
Second lactation in Denmark
Productive
life
Dairy form
0.51*
0.43*
0.51*
–0.42*
–0.53*
–0.33*
0.29*
0.37*
–0.50*
–0.42*
0.36*
0.35*
–0.34*
–0.17
0.36*
0.41*
–0.33*
–0.10
1Correlations between US productive life and Danish traits are
based on data from 104 bulls. The correlation between US productive life and the Swedish trait is based on data from 84 bulls.
Correlations between US dairy form and Danish traits are based on
data from 88 bulls. The correlation between US dairy form and the
Swedish trait is based on data from 88 bulls. Edits were made to
include only bulls with an approximate minimum of 50 daughter
equivalents in the genetic evaluations for diseases and reliabilities
for US productive life of 0.60 or greater or reliabilities for US linear
type of 0.70 or greater. Because of scaling, higher values are
desirable for the Danish and Swedish traits.
*Correlations among sire evaluations were different from 0 ( P <
0.05).
the unadjusted PL. Inclusion of PL in the US breeding program in the absence of direct evaluations for
diseases will likely have a desirable impact on diseases other than mastitis.
Most genetic correlations among selected type
traits and reproductive diseases, foot and leg diseases, and digestive and metabolic diseases were
small and did not indicate important relationships
between type and the diseases in this study. However,
as with the case for the composite disease traits, the
correlations between dairy form and the three disease
categories (reproductive diseases, foot and leg diseases, and digestive and metabolic diseases) were all
negative (range –0.34 to –0.64). It is interesting to
note that correlations of rump width, body depth,
strength, and stature with reproductive diseases, foot
and leg diseases, and digestive and metabolic diseases
also tend to be negative or near zero. Selection for
high dairy form scores may substantially increase
diseases other than mastitis in dairy cattle. In addition, selection for wider rumps, deeper bodies, more
strength, and taller cows will likely have little posiJournal of Dairy Science Vol. 82, No. 6, 1999
tive impact on disease resistance. In this study, there
was no indication that genetically larger body size or
more strength would be advantageous from a disease
perspective. These results indicate that selection for
smaller body size and increased yield to improve feed
efficiency will not likely increase disease incidence.
One might speculate that selection for smaller body
size might result in more metabolic stress and increased metabolic diseases, but these results do not
support such speculation.
Genetic correlations between disease traits and
dairy form calculated from residual correlations after
adjustment for PTA milk yield (Table 4 ) are all
negative and tend to be moderate in magnitude
(range –0.10 to –0.53). These results indicate that
dairy form has a significant association with diseases
other than mastitis, especially reproductive diseases,
that is independent of milk yield. The results of this
study are especially troublesome because dairy form
or a closely related trait is often utilized to choose
cows as bull dams. Often, cows with higher dairy form
scores are selected as bull dams when most other
available criteria are similar. Dairy form scores are
essentially utilized to confirm the accuracy of
recorded production. In addition, selection for higher
type scores essentially places a small direct emphasis
on increased angularity. Based on these results, selection for higher dairy form may be indirectly selecting
for cows that are more prone to reproductive diseases,
foot and leg diseases, and digestive and metabolic
diseases. Higher dairy form scores may be a reflection
of subclinical metabolic disease and selection for
higher dairy form scores may increase associated disease frequency. Perhaps dairy form could be used as a
marker for improved disease resistance, but the direction of selection would be opposite of what is currently
practiced. Selection for lower dairy form scores would
only be reasonable if intense selection for recorded
milk yield was practiced because dairy form and milk
yield are correlated. The current direct selection for
milk yield and simultaneous indirect selection for
yield based on dairy form is likely counter to improving disease resistance in dairy populations. Measured
milk, protein, and fat yields probably should be the
only traits used for improving yields; correlated body
traits should be used with caution.
The quadratic regression of reproductive diseases
in second lactation from Denmark on dairy form was
the only significant ( P < 0.05) quadratic regression
found in these analyses (Figure 1). Although not
significant ( P = 0.15), the quadratic regression of
reproductive diseases in first lactation from Denmark
on dairy form was similar in shape. Linear regression
1337
TRAITS CORRELATED WITH HEALTH
quadratic genetic relationships among the diseases
and the selected type traits were not important.
Although intermediate optimums may exist at the
phenotypic level for some type traits in relation to
disease incidence, we found little evidence in this
study to indicate that breeding programs should
select or mate to produce animals with intermediate
genetic values for the type traits that reflect locomotion or body characteristics.
CONCLUSIONS
Figure 1. Regression of Danish sire evaluations for reproductive
diseases in second lactation on US sire evaluations for dairy form.
Lines represent linear ( . . . ) and quadratic regressions ( – ) and are
based on data from 88 bulls with Danish and US genetic evaluations. Edits were made to include only bulls with an approximate
minimum of 50 daughter equivalents in Danish disease evaluations
and reliabilities for US type of 0.70 or greater. Danish sire evaluations had a mean of –0.22 and the standard deviation of 0.17 for
these 88 bulls (range from –0.80 to 0.16). The US genetic evaluations for dairy form had a mean of 0.44, and the standard deviation
was 1.08 for these 88 bulls (range from –2.23 to 3.04).
of reproductive diseases in second lactation in Denmark on dairy form had an intercept of –0.20 and a
linear coefficient of –0.066 (SE = 0.015). The quadratic regression of reproductive diseases in second
lactation in Denmark on dairy form had an intercept
of –0.17, a linear coefficient of –0.048 (SE = 0.017),
and a quadratic coefficient of –0.024 (SE = 0.011).
The quadratic relationship between reproductive diseases in Denmark and dairy form indicated that as
dairy form scores increased, reproductive diseases
tended to increase at an increasing rate. Those bulls
with very high dairy form scores had daughters with
the most reproductive disease problems.
Correlations among foot angle and reproductive
diseases, foot and leg diseases, and digestive and
metabolic diseases tended to be very low, which does
not support other reports ( 8 ) . Other studies (7, 8 )
have found important relationships among foot traits
and various measures of health. Correlations between
rump angle and foot and leg diseases were moderate
and positive. More slope from hooks to pins was
favorably associated with foot and leg diseases. Dutch
workers ( 2 ) found that more slope was also associated with decreased calving difficulty. Perhaps
selection for more slope to the rump can be justified
for maternal calving ease and for improved foot and
leg health.
With the exception of the quadratic regression of
reproductive diseases from Denmark on dairy form,
A substantial antagonistic genetic relationship exists between protein yield and diseases other than
mastitis. A desirable genetic correlation exists between US PL and diseases other than mastitis (especially if the effect of milk yield on PL and disease is
removed). Selection for increased PL may have a
desirable impact on the frequency of diseases other
than mastitis in US Holsteins.
Results from this study indicate that the genetic
correlations between most type traits and diseases
other than mastitis are small. Traits that are exceptions to this general result include dairy form and
rump angle. In this study, more slope from hooks to
pins was genetically associated with less frequent foot
and leg diseases. Likely more important, the genetic
correlations between dairy form and diseases were
unfavorable. Selection for higher dairy form scores
may substantially increase diseases other than mastitis and may compound the undesirable response in
diseases other than mastitis that accompanies selection for increased yield. Current practices of using
dairy form or equivalent information for final decisions in bull dam selection or in sire selection should
be reevaluated.
ACKNOWLEDGMENTS
Appreciation is expressed to 21st Century Genetics
(Shawono, WI), American Jersey Cattle Association
(Reynoldsburg, OH), Atlantic Breeders Cooperative
(Lancaster, PA), Holstein Association (Brattleboro,
VT), Select Sires, Inc. (Plain City, OH), and Sire
Power, Inc. (Tunkhannock, PA) for financial support.
The authors thank Mats Gundel from the Swedish
Association for Livestock Breeding and Production for
providing data. In addition, Jan Philipsson is thanked
for providing the sabbatical opportunity for G. W.
Rogers.
REFERENCES
1 Calo, L. L., R. E. McDowell, L. D. Van Vleck, and P. D. Miller.
1973. Genetic aspects of beef production among HolsteinJournal of Dairy Science Vol. 82, No. 6, 1999
1338
ROGERS ET AL.
Friesian pedigree selected for milk production. J. Anim. Sci. 37:
676–682.
2 De Jong, G. 1991. What is the optimal rump angle for the dairy
cow? Veepro Holland 11:20–21.
3 INTERBULL. 1996. Sire evaluation procedures for non-dairyproduction and growth and beef production traits practiced in
various countries. Pages 129–139 and 153–158 in INTERBULL
Bull. No. 13. Int. Bull Eval. Serv., Uppsala, Sweden.
4 Koenen, E.P.C., and A. F. Groen. 1998. Genetic evaluation of
body weight of lactating Holstein heifers using body measurements and conformation traits. J. Dairy Sci. 81:1709–1713.
5 Lin, H. K., P. A. Oltenacu, L. D. Van Vleck, H. N. Erb, and R. D.
Smith. 1989. Heritabilities of and genetic correlations among
six health problems in Holstein cows. J. Dairy Sci. 72:180–186.
6 Lyons, D. T., A. E. Freeman, and A. L. Kuck. 1991. Genetics of
health traits in Holstein cattle. J. Dairy Sci. 74:1092–1100.
7 Rogers, G. W. 1994. Requirements and uses of evaluations for
health and reproductive traits. Proc. 5th World Congr. Genet.
Appl. Livest. Prod. 17:81–88.
8 Rogers, G. W. 1996. Using type for improving health of the
udder and feet and legs. Pages 33–41 in Proc. Int. Workshop
Journal of Dairy Science Vol. 82, No. 6, 1999
Genet. Improvement Functional Traits Cattle. INTERBULL
Bull. No. 12. Int. Bull Eval. Serv., Uppsala, Sweden.
9 Sander-Nielsen, U., G. A. Pedersen, J. Pedersen, and J. Jensen.
1997. Genetic correlations among health traits in different lactations. Pages 68–77 in Proc. Int. Workshop Genet. Improvement Functional Traits Cattle; Health. INTERBULL Bull. No.
15. Int. Bull Eval. Serv., Uppsala, Sweden.
10 Shook, G. E. 1989. Selection for disease resistance. J. Dairy Sci.
72:1349–1362.
11 VanRaden, P. M., and E.J.H. Klaaskate. 1993. Genetic evaluations of length of productive life including predicted longevity of
live cows. J. Dairy Sci. 76:2758–2764.
12 VanRaden, P. M., and G. R. Wiggans. 1991. Derivation, calculation, and use of national animal model information. J. Dairy
Sci. 74:2737–2746.
13 Veerkamp, R. F. 1998. Selection for economic efficiency of dairy
cattle using information on live weight and feed intake: a
review. J. Dairy Sci. 81:1109–1119.
14 Weigel, K. A., T. J. Lawlor, Jr., P. M. VanRaden, and G. R.
Wiggans. 1998. Use of linear type and production data to supplement early predicted transmitting abilities for productive
life. J. Dairy Sci. 81:2040–2044.